KCSE Topic Predictor AI

📋 Overview

KCSE Topic Predictor AI analyzes 20+ years of KCSE past papers to predict likely exam topics. The system uses machine learning to identify patterns in question distribution across subjects.

🚀 Features

• PDF Processing: Extracts questions from past papers (2000-2024)
• AI Predictions: Machine learning model predicts topic probabilities
• Subject Filtering: Filter predictions by specific papers
• Confidence Scores: Visual indicators of prediction reliability

🛠️ Tech Stack

• Backend: Django, Django REST Framework
• Frontend: React, Tailwind CSS
• ML: scikit-learn, TF-IDF Vectorization
• Database: SQLite/PostgreSQL
• PDF Processing: pdfplumber, pytesseract

🖥️ Usage

🔍 View Predictions

⚡ Quick Start

1. Backend Setup

cd backend
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
python manage.py migrate
python manage.py runserver

2. Frontend Setup

cd frontend
npm install
npm run dev

Subject Support

• Mathematics (MATH)
• Chemistry (CHEM)
• Physics (PHYS)
• Biology (BIO)
• English (ENG)

Subject Support

1. Extract: PDFs are processed to identify questions and topics
2. Analyze: Historical data is analyzed for topic frequency patterns
3. Predict: ML models predict topic probabilities for upcoming exams
4. Display: Results show confidence scores and paper distributions